基于资源整合的节能虚拟网络重配置算法

Energy-aware virtual network reconfiguration algorithm based on resource consolidation

针对虚拟网络映射中能耗过高、接收率偏低和负载不够均衡等问题,提出一种基于虚拟资源整合的综合性重配置算法——HEAR算法。该重配置算法分为两个阶段:节点重配置阶段优先将映射虚拟节点最少的物理节点上的虚拟节点及其相连虚拟链路迁移,挂起或关闭空负载的物理节点来达到节能的目的;此外对这些迁移节点的目标物理节点进行筛选,避免选择过度拥塞的物理节点达到提高接收率和均衡负载的目的。链路重配置阶段采用能耗感知的方法选择可用于迁移的物理链路集合,再用Dijkstra算法选择最短物理路径并将相关路径迁移过去。实验结果表明,HEAR算法比启发式重配置算法平均能耗下降约20%,接收率提高约10%。

Concerning the high energy consumption, low acceptance rate and unbalanced load in virtual network embedding,a comprehensive energy-aware virtual network reconfiguration algorithm based on resource consolidation,namely HEAR algorithm,was proposed,which consists of two stages including node reconfiguration and link reconfiguration. In node reconfiguration stage,the virtual nodes on the physical node with least mapping virtual nodes and their relevant virtual links were moved to other physical nodes except congested nodes to improve acceptance rate and load balance,as well as suspending or closing the physical nodes with empty load to save energy. In link reconfiguration stage,the energy-aware method was adopted to select substrate link candidate set for migration,and Dijkstra algorithm was used to select the shortest available physical path to redeploy the virtual links on it. The simulation results show that,compared with energy-aware relocation heuristic algorithm,HEAR algorithm can reduce energy consumption by about 20%,and increase acceptance rate by about10%,which means it can save energy consumption,improve the acceptance rate.